Machine Learning Algorithm for Predicting Distant Metastasis of T1 and T2 Gallbladder Cancer Based on SEER Database

Area of Science:

• Oncology
• Medical Informatics
• Machine Learning

Background:

• Gallbladder cancer (GBC) poses a significant challenge, particularly regarding distant metastasis (DM) in early stages.
• Accurate prediction of DM is crucial for effective treatment planning and improved patient outcomes.

Purpose of the Study:

• To develop and validate a machine learning (ML) model for forecasting the risk of distant metastasis (DM) in patients with T1 and T2 gallbladder cancer (GBC).
• To identify independent risk factors associated with DM in early-stage GBC.
• To create a predictive tool for clinical use.

Main Methods:

• Utilized data from 4371 patients with T1 and T2 GBC from the SEER database (2004-2015).
• Applied and evaluated seven ML algorithms, including logistic regression (LR) and Random Forest (RF).
• Assessed model performance using metrics like the area under the receiver operating characteristic curve (AUC).

Main Results:

• Out of 4371 patients, 764 (17.4%) developed distant metastasis.
• The Random Forest (RF) model demonstrated superior predictive performance compared to other algorithms.
• A nomogram was developed to predict DM in early T-stage GBC patients.

Conclusions:

• The Random Forest (RF) algorithm achieves high accuracy in predicting distant metastasis (DM) for gallbladder cancer (GBC) patients.
• This ML-driven approach can enhance diagnostic accuracy for clinicians.
• The developed RF model and web calculator can optimize treatment strategies and improve patient outcomes.